The HIV envelope glycoprotein is widely believed to be an important determinant of viral fitness and transmissibility. How Env affects transmission has not been carefully defined. In our preliminary work, we studied viruses in a unique cohort of transmission pairs in which the source virus was confirmed with phylogenetic sequencing, and compared these viruses to those found in a cohort of non-transmitting pairs. Using a pseudotyped viruses and a single-cycle assay, we found that R5 viruses from subjects who transmit virus appear to have greater efficiency in using CCR5 compared to R5 viruses from subjects who do not transmit virus. This apparent link between variability in CCR5 utilization efficiency and successful virus transmission was not explained by the level of plasma viremia. In this proposed project we will test the hypothesis that the capacity of HIV to efficiently utilize CCR5 is a strong and independent predictor or transmissibility. In Specific Aim 1, we will first attempt to confirm the association between R5 utilization efficiency and transmissibility using pseudotyped viruses from a large number of transmitting and non-transmitting individuals infected with diverse HIV subtypes. We will then examine the specific virus-related properties related to this interaction (e.g., binding affinity, fusogenicity, fitness/infectivity, sensitivity to neutralizing antibodies and other inhibitors including chemokines and R5 inhibitors, the number and location of N-linked glycosylation sites or variable loop lengths). In Specific Aim 2 we will determine host factors in the source partner which are associated with harboring more transmissible HIV, focusing on the CCR5 receptor. Finally, in Specific Aim 3, we will investigate the influence of these host factors on the evolution of HIV transmissibility. These broadly defined objectiveswill be addressed in an expanded cohort of xxx transmitting pairs and yyy non-transmitting pairs in San Francisco. In order to investigate the role of virus subtype, we will extend our work to southern Brazil, where there has been a recent sharp increase in the prevalence of subtype C among newly infected individuals. A total of 50 transmission and 50 non-transmitting pairs will be identified, of whom approximately 50% are expected to have harbor subtype C virus.